HAL Id: hal-02627459
https://hal.inrae.fr/hal-02627459
Submitted on 26 May 2020
HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.
Distributed under a Creative Commons Attribution| 4.0 International License
Richard Whittington, Karsten Donat, Maarten F. Weber, David Kelton, Søren Saxmose Nielsen, Suzanne Eisenberg, Norma Arrigoni, Ramón Juste,
Jose Luis Sáez, Navneet Dhand, et al.
To cite this version:
Richard Whittington, Karsten Donat, Maarten F. Weber, David Kelton, Søren Saxmose Nielsen, et
al.. Control of paratuberculosis: who, why and how. A review of 48 countries. BMC Veterinary
Research, BioMed Central, 2019, 15 (1), �10.1186/s12917-019-1943-4�. �hal-02627459�
R E V I E W Open Access
Control of paratuberculosis: who, why and how. A review of 48 countries
Richard Whittington
1*, Karsten Donat
2,3, Maarten F. Weber
4, David Kelton
5, Søren Saxmose Nielsen
6, Suzanne Eisenberg
7, Norma Arrigoni
8, Ramon Juste
9, Jose Luis Sáez
10, Navneet Dhand
1, Annalisa Santi
8, Anita Michel
11, Herman Barkema
12, Petr Kralik
13, Polychronis Kostoulas
14, Lorna Citer
15, Frank Griffin
16, Rob Barwell
17, Maria Aparecida Scatamburlo Moreira
18, Iva Slana
13, Heike Koehler
19, Shoor Vir Singh
20, Han Sang Yoo
21, Gilberto Chávez-Gris
22, Amador Goodridge
23, Matjaz Ocepek
24, Joseba Garrido
25,
Karen Stevenson
26, Mike Collins
27, Bernardo Alonso
28, Karina Cirone
29, Fernando Paolicchi
29, Lawrence Gavey
30, Md Tanvir Rahman
31, Emmanuelle de Marchin
32, Willem Van Praet
33, Cathy Bauman
5, Gilles Fecteau
34,
Shawn McKenna
35, Miguel Salgado
36, Jorge Fernández-Silva
37, Radka Dziedzinska
13, Gustavo Echeverría
38, Jaana Seppänen
39, Virginie Thibault
40, Vala Fridriksdottir
41, Abdolah Derakhshandeh
42, Masoud Haghkhah
42, Luigi Ruocco
43, Satoko Kawaji
44, Eiichi Momotani
45, Cord Heuer
46, Solis Norton
47, Simeon Cadmus
48,
Angelika Agdestein
49, Annette Kampen
49, Joanna Szteyn
50, Jenny Frössling
51, Ebba Schwan
52, George Caldow
53, Sam Strain
54, Mike Carter
55, Scott Wells
56, Musso Munyeme
57, Robert Wolf
58, Ratna Gurung
59, Cristobal Verdugo
36, Christine Fourichon
60, Takehisa Yamamoto
44, Sharada Thapaliya
61, Elena Di Labio
62, Monaya Ekgatat
63,
Andres Gil
64, Alvaro Nuñez Alesandre
65, José Piaggio
64, Alejandra Suanes
66and Jacobus H. de Waard
67Abstract
Paratuberculosis, a chronic disease affecting ruminant livestock, is caused by Mycobacterium avium subsp.
paratuberculosis (MAP). It has direct and indirect economic costs, impacts animal welfare and arouses public health concerns. In a survey of 48 countries we found paratuberculosis to be very common in livestock. In about half the countries more than 20% of herds and flocks were infected with MAP. Most countries had large ruminant
populations (millions), several types of farmed ruminants, multiple husbandry systems and tens of thousands of individual farms, creating challenges for disease control. In addition, numerous species of free-living wildlife were infected. Paratuberculosis was notifiable in most countries, but formal control programs were present in only 22 countries. Generally, these were the more highly developed countries with advanced veterinary services. Of the countries without a formal control program for paratuberculosis, 76% were in South and Central America, Asia and Africa while 20% were in Europe. Control programs were justified most commonly on animal health grounds, but protecting market access and public health were other factors. Prevalence reduction was the major objective in most countries, but Norway and Sweden aimed to eradicate the disease, so surveillance and response were their major objectives. Government funding was involved in about two thirds of countries, but operations tended to be funded by farmers and their organizations and not by government alone. The majority of countries (60%) had voluntary control programs. Generally, programs were supported by incentives for joining, financial compensation and/or penalties for non-participation. Performance indicators, structure, leadership, practices and tools used in control programs are also presented. Securing funding for long-term control activities was a widespread problem.
(Continued on next page)
© The Author(s). 2019Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
* Correspondence:richard.whittington@sydney.edu.au
1School of Veterinary Science, Faculty of Science, University of Sydney, 425 Werombi Road, Camden, NSW 2570, Australia
Full list of author information is available at the end of the article
(Continued from previous page)
Control programs were reported to be successful in 16 (73%) of the 22 countries. Recommendations are made for future control programs, including a primary goal of establishing an international code for paratuberculosis, leading to universal acknowledgment of the principles and methods of control in relation to endemic and transboundary disease. An holistic approach across all ruminant livestock industries and long-term commitment is required for control of paratuberculosis.
Keywords: Paratuberculosis, Control, Review, Prevalence, Cattle, Sheep, Goat, Camelid, Deer, Wildlife
Background
Paratuberculosis is a mycobacterial disease of ruminants caused by Mycobacterium avium subsp. paratuberculosis (MAP) . It begins as a localised infection that may become systemic and often results in chronic granulomatous en- teritis leading eventually to weight loss, (diarrhoea in some species) and death. This syndrome is also known as Johne ’ s disease. Depending partly on how long it has been present in a herd, it can manifest as isolated clinical cases or persistent outbreaks. The pathogenesis of paratubercu- losis is similar to tuberculosis and other mycobacterial dis- eases, MAP being a specialised intracellular pathogen that triggers a debilitating immunopathological response, sub- clinical infections however being more common [1]. Most species of ruminant livestock are susceptible to paratuber- culosis. It is present in many countries and is perceived to be an important disease for reasons including its impacts on the economy, the impact of clinical disease on animal welfare, and public health.
Why should it be controlled?
Economic impacts
Economic losses due to paratuberculosis are a key driver to control MAP. While the economic losses in dairy cat- tle have been extensively studied, the difficulties in quantifying them have been highlighted in two reviews [2, 3]. Losses to the dairy farmer consist of losses before, during or after culling [4]. Losses before culling may in- clude reduced milk production of variable magnitude [5–
10], increased somatic cell counts [7, 8, 10–12], increased incidence of clinical mastitis [10, 13, 14], reduced fertility [2, 15, 16], increased susceptibility to other diseases [17, 18] and costs of testing and treatment [4]. Cattle infected with MAP have higher on-farm mortality and cull rates [19–23], as do veal calves that originated from dairy herds with paratuberculosis [24]. Because MAP infection is asso- ciated with a lower body weight [25, 26], the slaughter value of infected cattle is reduced [4, 19, 20, 27, 28]. Costs after culling are the loss of unrealised future income by culling an individual [4, 29, 30].
The total annual economic losses per cow in infected USA dairy herds were estimated at US$21 to US$79 [20, 31–34] while those in infected Australian, Canadian, French, and UK dairy herds were estimated at A$45- A$88 [29, 35], CDN$49 [36, 37], €234 [38] and GBP27
[39], respectively. In ‘ average ’ Dutch dairy herds (both infected and uninfected), these losses were estimated at up to € 67 per cow per year [32]. The effect of the reduction in milk production caused by MAP infec- tion in US dairy herds to the US economy was esti- mated to be an annual loss of US$200 million ± US$
160 million [40].
In beef cattle, the losses due to MAP infections de- pend on the markets targeted by the farmer. Pedigree farmers selling breeder cattle are much more affected than farmers selling all their cattle directly for slaughter or to feedlots [41]. As with dairy cattle, test-positivity in beef cattle was associated with lower cow fertility, weight loss, and lower calf body weight at birth and at 205 days of age [42]. Total annual losses per cow in the first 15 – 20 years after MAP introduction in suckler herds were estimated to be on average GBP16 in British herds [43]
and € 40 in French herds [38], whereas in Dutch suckler cow herds (both infected and uninfected) they were € 10 (small herds) to € 28 (large herds) [44]. Farmers and vet- erinarians perceived the annual losses per cow in in- fected suckler cow herds to be around on average US$16 to US$17 [45]. However, most of the potential losses due to MAP infection are hidden or opportunity costs, not out-of-pocket losses [46].
Published estimates of economic losses in other do- mestic ruminant species are sparse. In an Australian study of 12 sheep flocks, the average annual mortality rate due to paratuberculosis varied between 6.2 and 7.8%, resulting in a decrease of the average gross margin of between 6.4 and 8.5% [47]. However, in some Austra- lian flocks, mortality approached 20% per annum [48].
In addition to mortalities, weight loss in affected sheep
is of economic importance [49]. Infected fine wool Me-
rino flocks in New Zealand lost US$1.5 per ewe annually
which was about 3-fold as much as in mutton breeds,
while a case fatality rate of 1.2 – 2.7% was the main cost
factor (Gautam et al. 2019, in press). The economic
losses of paratuberculosis to the British sheep industry
were estimated at GBP 0.4 to 32 million per annum, de-
pending on assumptions on the prevalence of infection,
the mortality rate and the replacement strategy [50]. A
recent economic study showed that MAP infection de-
creased the profit efficiency from 84 to 64% in Italian
dairy sheep and goat farms [51].
In addition to the economic production losses de- scribed above, indirect losses may arise from trade re- strictions at both the international [52] and national levels. For instance, at the national level, Dutch dairy herds can only deliver milk to their milk processors if they periodically test their herd and cull any test-positive cattle [53]. Furthermore, it is important to realise that production losses caused by MAP not only impact in- fected farms, but also have a negative economic impact on consumers [54]. Despite the magnitude of these eco- nomic losses to both infected farms and consumers, probably the most important economic driver to control MAP infection in livestock is uncertainty about the po- tential involvement of MAP in human disease [55, 56].
Even in the absence of a proven link between MAP and human disease, the economic consequences of reduced milk demand could be large if consumers’ perception of risk is large or if risk-mitigation strategies were per- ceived to be ineffective [57].
Potential public health implications There are myriad publications concerning the link between MAP and dis- eases in humans. The authors of a series of review arti- cles with rigorous methodologies concluded that sources of human exposure existed [58, 59], and “while the zoo- notic potential of M. paratuberculosis cannot be ignored, due to important knowledge gaps in understanding its role and importance in the development or progression of human disease, its impact on public health cannot yet be quantified or described” [60] . Thus, MAP can infect humans but the infection may or may not cause Crohn’s disease, the condition with which MAP is most often linked. Consequently, the reviewers concluded that steps beyond the programs that the dairy and ruminant indus- tries had already developed for animal health and eco- nomic reasons could not be justified by public health authorities [60]. In other words, there was reliance on animal health authorities to take the lead in managing the risk, i.e. to reduce the exposure of humans to MAP by controlling paratuberculosis in livestock.
Trade of livestock A major conundrum exists regarding trade of livestock and their products. Whereas members of the World Trade Organization (WTO) are required to take an “ appropriate ” , scientifically based level of sanitary measures to protect human and animal health, WTO members cannot discriminate between members where similar conditions prevail, e.g. a WTO member cannot require freedom of MAP in traded livestock if they themselves do not carry out activities to document such freedom [52]. The World Organization for Animal Health (OIE), whose standards are recognised by WTO, offers no guidance on paratuberculosis [61]. The Inter- national Association for Paratuberculosis therefore
provided guidelines for movement of livestock, accord- ing to the rules laid down in the Sanitary and Phytosani- tary (SPS) Agreement of WTO [52]. Furthermore, the parliament of the European Union (EU) in 2016 decided to list diseases according to specific criteria through the EU Animal Health Law [62]. This law should be imple- mented in all member states of the EU in 2021 with spe- cific rules for each disease.
The European Commission in 2016 asked the Euro- pean Food Safety Authority (EFSA) for advice regarding paratuberculosis, and EFSA proposed that paratubercu- losis complied with the criteria to be categorized with regards to movement control (Category D) and surveil- lance (Category E) to avoid spread of MAP [56]. After the exchange of views between European Commission and Member States in the Standing Committee on Plants, Animals, Food and Feed, paratuberculosis was fi- nally included only under Category E (surveillance) for the listed animal species Bison spp., Bos spp., Bubalus spp., Ovis spp., Capra spp., Camelidae and Cervidae.
Category D and other categories were disregarded, mainly due to the lack of individual animal sensitivity of current diagnostic tests and the difficulties associated with granting paratuberculosis free status to herds, areas and countries. Somewhat paradoxically the same tests are being used within countries for disease control pur- poses (see below).
Epidemiology and pathogenesis of paratuberculosis
MAP infections occur. in domestic and wild ruminants throughout the world, with some monogastric species also having been reported as infected [56, 63]. Apparent spillover or endemicity has been described in wildlife in some countries, with economic consequences for farmed wildlife and unknown consequences for conservation of free-ranging wildlife, endangered species or those in cap- tive breeding programs [64, 65]. Transmission of MAP is often insidious [66]. No country can be considered without risk of introduction of the disease. Thus, free- dom from disease is challenging to document and re- quires extensive surveillance and reporting. Sweden claims to be free of paratuberculosis in cattle since 2008, and Norway has had no known cases of paratuberculosis since 2015.
The age-profile within herds and flocks is a factor in-
fluencing the occurrence of clinical disease because of
the long incubation period. While the onset of clinical
paratuberculosis is usually at adult ages, many individ-
uals become infected during the first weeks of life when
they are exposed to MAP from older animals shedding
the bacterium in their faeces. Infection at older ages is
also possible [67, 68]. MAP is mainly transmitted dir-
ectly by the faecal-oral route, including via faecal con-
tamination of the udder or pasture, but the formation of
aerosols might take bacteria to environmental niches from which it could spread to naive cattle [69]. Other routes like in utero are not negligible [70]. Lower envir- onmental exposures may explain why the disease is much less prevalent in extensively raised beef cattle than in intensively raised dairy cattle. Commonly in beef cat- tle herds the animals are slaughtered before they reach the more advanced stages of infection where heavy MAP shedding occurs; thus there is less opportunity for envir- onmental contamination. However, this does not explain the widespread disease in small ruminants raised in ex- tensive conditions like those in Australia or Spain. MAP is resistant to acidic soils and low temperatures, includ- ing freezing, but it seems to be less resistant in hot and dry climates. Protected from sunlight, MAP survived for up to 55 weeks in a dry fully shaded environment [71]. If vegetation is removed and shading is limited, survival is reduced to a few weeks [72].
Within species, some breeds of livestock are more likely than others to develop clinical paratuberculosis in a given time frame [73]. Further, associations between susceptibility to paratuberculosis and specific genetic markers have been demonstrated [74], notwithstanding the problems of inconsistent phenotypic classification [75, 76]. Based on histopathological evidence, latent forms of the disease, probably representing varying de- grees of resistance, account for the majority of all infec- tions [77].
After oral exposure, the first site of MAP colonization is the ileal and jejunal Peyer’s patch, an organized lymphoid tissue in the intestinal mucosa and submucosa, from where it spreads to the mucosal lamina propria [78]. MAP shows a clear tropism for this site, exiting the host via the intestinal lumen, thus explaining the main route of spread in faeces [79]. The amount of mycobac- teria shed in faeces is most relevant for transmission of the infection through environmental contamination.
Practices and tools for the control of paratuberculosis
The introduction of MAP into a herd is often recognized only after spread has occurred [66]. Control of paratu- berculosis then depends on population-level measures such as the culling of animals that are shedding MAP, applying hygienic measures aimed at reducing infection of neonatal/young stock, and vaccination. Epidemio- logical models in dairy cattle suggest that test and cull or actions targeting infection routes are effective strat- egies to decrease MAP prevalence [80–82]. These results are reflected in independent data from other herds [83, 84]. Another risk factor that can be managed is the introduction of livestock onto the farm [83, 85]. How- ever, based on studies conducted in several countries the compliance of farmers with recommendations on con- trol of paratuberculosis can be low [86, 87].
MAP control programs can have different goals ran- ging from reduction of clinical cases and/or MAP preva- lence in a herd, which has been shown to be feasible in general [84, 88, 89], to eradication of MAP from the herd, which can be achieved only in some herds [90, 91].
Stamping out can be mandated when the prevalence is low to zero.
Depending on the goal, several control strategies apply. One is the identification and elimination of clinic- ally diseased and/or subclinically infected animals (‘test- and-cull’) [92]. However, as a stand-alone measure this will not eradicate paratuberculosis in the long term [93–
95]. A second strategy applicable to all species is the prevention of MAP transmission within the herd by en- hancing on-farm biosecurity especially during rearing of young stock. The main intervention strategy in dairy cat- tle operations is preventing calves from contacting the faeces of adult cows [96]. This avoids faecal-oral trans- mission and in practice is achieved by improving calving area hygiene and management of colostrum/milk feed- ing. A process called “snatching”, that involves removing goat kids at birth and separating them from their dams, was successfully used to sanitize goat herds in previously endemic areas in Norway [97], and is used in the Netherlands. Basic preventive measures concentrate on improving hygiene measures in young stock raising [98]
but should be extended to all management areas to in- crease success [32, 88]. Small-scale field trials as well as simulation studies have shown that the most effective con- trol strategy consists of a combination of ‘ test-and-cull ’ and increasing on-farm biosecurity [94, 99, 100]. The re- sults of a modelling study indicated that improving calf hygiene and ‘ test-and-cull ’ were both necessary to stabilize the herd status, but reduced calf exposure was confirmed to be the most influential measure, followed by testing fre- quency and the proportion of infected animals that were detected and subsequently culled [101]. Culling the pro- geny of known infected cows may also be considered as part of the control strategy as there is a significant rate of in utero infection with MAP [70]. Considering the finite environmental survival of MAP, pasture and grazing man- agement can be utilised to reduce the exposure of exten- sively grazed livestock [71, 72].
As a further tool, biosecurity measures have to be in place to control the transmission between herds and to protect MAP-negative herds from new infection or re- introduction of MAP. The purchase of sub-clinically in- fected animals is considered to be the main factor for between-herd transmission [102]. Therefore, MAP-herd status certification that reflects the risk of MAP trans- mission by animals originating from that herd is essen- tial [103, 104].
Vaccination of ruminants has been demonstrated in
controlled studies to be effective in reducing the clinical
incidence of paratuberculosis, delaying the onset of the disease and reducing faecal shedding of MAP, thus redu- cing the economic losses as well as transmission of MAP [105, 106]. There can be considerable benefit of vaccination relative to the cost [107] but there does not appear to have been a critical review of this topic. Vac- cination is not widely used in paratuberculosis control in cattle. This is because of the risk of interference with intradermal testing for bovine tuberculosis [108–110].
Importantly and in contrast, in the Australian, Icelandic and Spanish sheep industries, vaccination has been a key element in control leading to a significant reduction in within-herd prevalence [111, 112].
A very important element in MAP-control is commu- nication to farmers and veterinarians about the import- ance and the components of the program [113–115] and the need to consider farmers’ attitudes towards the im- plementation of control measures on their farms [116].
Diagnostic tests for paratuberculosis
Diagnostic tests used to detect paratuberculosis are gen- erally imperfect yet most are useful if applied properly when a specific purpose has been identified. Examples of purposes are establishment of populations at low risk of infection, reduction of economic impact in diseased ani- mals, eradication, confirmation of clinical disease, sur- veillance and prevalence estimation [117]. The purpose defines the stage of disease that must be detected: “in- fected” (MAP has colonised the animal’s tissues), “infec- tious” (MAP is being shed in faeces) or “affected by MAP” (the animal is clinically diseased) [118]. Detection should result in a decision [119] which could, for ex- ample, be culling an animal to reduce the transmission of MAP or to reduce its suffering. The relevant stage of the disease to target and the applicable diagnostic test are linked. Case definitions and terminology are very im- portant for consistent application of tests and interpret- ation of test results [76].
A wide range of tests exist. In general, they can be classified as tests for MAP, or tests for the host immune response against MAP. Bacterial culture or polymerase chain reaction (PCR) used for faecal samples are direct measures of bacterial shedding of the individual, whereas enzyme-linked immunosorbent assay (ELISA) used for serum or milk, and agar gel immunodiffusion (AGID) and complement fixation test (CFT) are measures of humoral immune responses, also of the individual. Tests for the humoral immune response are applicable in more advanced stages of the pathogenesis associated with major bacterial shedding, reduced milk yield and reduced weight at slaughter [27, 120–122]. Immune re- sponses detected by such tests are not always specific for MAP infection [123], and similarly faecal tests may
detect MAP that has merely been ingested and then pas- sively shed, i.e. without true infection [124].
Faecal culture and faecal PCR are widely used tests for MAP, while ELISA is commonly used to detect an anti- body response against MAP. AGID and CFT are less commonly used and less sensitive compared to ELISA [120, 125]. The choice between tests can be related to costs and logistics [100, 126]. Even though the sensitivity of serum ELISA is considerably lower than the sensitivity of faecal culture or faecal PCR, its sensitivity is fairly high in cattle in advanced stages of infection [118, 127]
and higher S/P readings are associated with a higher proportion of cattle shedding MAP in their faeces [128].
A definitive diagnosis of MAP infection in the individual may be made on gross and histopathological examin- ation, but even this test is not perfect and may require a high number of tissues to rule out infection [129].
The accuracy of diagnosis of the individual can be in- creased significantly if the results for the individual are used in context of historical results of the herd or flock of origin [130, 131]. These results can include individual diagnostics as mentioned above, or herd- or flock-level diagnostics such as culture or PCR on environmental samples [132] [133], or use of pooling [90, 126, 134, 135]. Herd- and flock level diagnostics can also be used to inform risk mitigation strategies for the herd or flock.
If there are no historical results, different types of tests can be used in parallel or in series to increase the sensi- tivity or specificity, respectively.
The Ziehl-Neelsen (ZN) stain of faecal or tissue smears may serve as a fast diagnostic tool to support clinical diagnosis, but serum ELISA is superior for this purpose [127, 136]. Skin testing to detect cell-mediated immune responses is also an option [137, 138], but its use may affect the skin test for Mycobacterium bovis . Furthermore, detection of cell-mediated immune re- sponses using the skin test or in vitro tests such as the interferon-gamma release assay may merely reflect ex- posure to MAP and not be related to infection status.
Lastly, bulk tank milk ELISA is also used on occasion, but low within-herd prevalence usually precludes a prac- tical application [139] even though there is statistical evidence of diagnostic value [140, 141]. OIE [117] pro- vides an overview of different tests for different pur- poses, but a multitude of considerations are needed to establish a useful test-strategy (see e.g. [142]).
History of paratuberculosis control programs
Control programs for paratuberculosis have existed since
early last century. Their chronology was documented by
Benedictus et al. [113]: the first programs were set up in
France in the 1920’s followed by the Netherlands in
1942, Great Britain and Norway in the 1960’s, then
Cyprus and the United States of America in the 1970s.
Iceland had a program to control paratuberculosis in sheep in the 1950s [112]. The Australian program began in the 1990’s [143]. Control programs for paratuberculo- sis were then implemented in other developed countries, and excellent reviews of these cover the period up until 2012, bringing information together particularly for cat- tle [144–147]. However, there are other important, sus- ceptible, livestock species and since 2012 new programs have commenced (e.g. Ireland, Italy) while others have been discontinued or substantially modified (e.g.
Australia, United States of America). While recent country-specific information may exist [148], in general there is a lack of authoritative information and it is cur- rently impossible to ascertain from any single source what is being done about paratuberculosis in different countries, or the reasons for coordinated action. Fur- thermore, it is difficult to find information on the global distribution of paratuberculosis. Consequently, animal health authorities in many countries are not in a good position to make recommendations to their own govern- ments or domestic animal industries, and sometimes they are forced to respond quickly to changing circum- stances with incomplete information.
Scope of this study
The aims of this study were to assess the existence and nature of control programs for paratuberculosis for the period 2012–2018 in countries for which there is reliable information. Further aims were to assess the rationale for control programs, outcomes of past and current con- trol programs, and to make general recommendations for future control programs.
Methods
ParticipantsCollaborators in different countries were identified using chain referral sampling commencing in November 2017 with an invitation to members of the governing board of the International Association for Paratuberculosis (IAP), a not for profit expert body with aims to promote know- ledge about paratuberculosis (http://www.paratuberculo- sis.net), to join a project team. Governing board members are elected by general members of the IAP in countries with multiple IAP members and hence were likely to have strong interest in paratuberculosis control and a network of expert contacts. Through a process of sequential referral from the board members, and subse- quently from those whom they nominated, the senior author progressively invited experts to participate as col- laborators. Criteria for nomination included: knowledge of, or access to, detailed information about paratubercu- losis and national and/or regional control programs in cattle, sheep, goats, camelids, deer, wildlife or zoological collections; approval from employer to participate (if
applicable); and willingness to work to a tight schedule in 2018. Those who agreed were sent additional infor- mation and asked to comment on the scope of the study and to complete a detailed on-line questionnaire, which was designed to collect information in a consistent man- ner. Participants ( n = 27) who were delegates to the 14th International Colloquium on Paratuberculosis, Riviera Maya, Mexico met face to face on June 6, 2018 to dis- cuss the project, review the results of preliminary ana- lysis and recommend topics for further exploration in this paper. The outcomes were circulated to all partici- pants for comment to inform the composition of this report.
Questionnaire
A draft questionnaire was developed using an on-line platform (Survey Monkey, San Mateo, A, USA) and pre- tested on ten volunteer collaborators each of whom had past experience with surveys for paratuberculosis, from ten countries (Australia, Canada, Czech Republic, Denmark, Germany, Ireland, the Netherlands, New Zea- land, South Africa and Spain). Their comments were used to improve the questionnaire, which was then made available by an e-mail link to all participants. A copy of the questionnaire is provided as Additional file 1.
Data were exported to Excel (Microsoft Corp., Redmond, WA). To enable country-level assessment when more than one questionnaire was completed for a given coun- try (typically where there were separate responses for different regions of a country), data were aggregated as follows. For animals: the largest regional population on the logarithmic scale was selected; numbers of farms were summed; the midpoint of average herd sizes and the overall minima and maxima were used. At country level: the disease was considered notifiable if it was at re- gional level; the highest prevalence estimate at regional level was selected; Yes was selected if this was a re- sponse in any region; the earliest year for all start events and the latest year for end events were selected; all re- sponses to open questions, and comments, from all re- gions were concatenated; other data in separate fields at regional level were brought to country level.
Statistical analyses
Summary statistics were calculated, and box-and- whisker plots were created for numeric variables, namely, number of farms and average, minimum and maximum herd sizes. The y-axes of box-plots were log
10transformed to improve their interpretability as their distributions were highly right skewed.
Frequency tables and bar charts were created for cat-
egorical variables such as animal population sizes, num-
ber of farms and herd/individual prevalence estimates
for each species. Maps were prepared in Mapline (ver- sion July 2018, www.mapline.com).
Source of herd and individual prevalence estimates for each species were classified into (a) objective, if the source was one of the following: apparent or true preva- lence based on serology, apparent or true prevalence based on individual or bulk milk ELISA, or apparent or true prevalence based on faecal culture or PCR; (b) sub- jective, if the source was one of the following: abattoir monitoring, other active surveillance, passive surveil- lance, or other; or (c) not applicable. Responses to open questions were manually paraphrased using consistent terms developed following assessment of all responses, summarised into groups and tabulated for analysis.
The relationship between herd-level prevalence esti- mates and log transformed herd size was determined using ordinal logistic regression. Number of farms and herd sizes were compared between countries with and without a paratuberculosis control program using Wil- coxon two sample tests. Statistical analyses were con- ducted using SAS (© 2002–2012 by SAS Institute Inc., USA), whereas graphs were produced using R Version 3.5.1 [149] and Excel. Proportions were compared using a two-sample two-sided binomial test in Genstat (VSN International).
Definitions Paratuberculosis
This was broadly defined as the disease (Johne’s disease), which has characteristic pathology and may be clinical, or infection with M. avium subsp. paratuberculosis (MAP) . The terminology used for paratuberculosis case definitions is available elsewhere [76].
Control program
The scope for a livestock disease control program in general was an ongoing process of measures intended to
interfere with the unrestrained occurrence of the disease based on Thrusfield [150]. The measures must include elements of planning, coordination, documentation and evaluation, and may be conducted locally, regionally, or nationally. Examples of measures included: estimation of prevalence to inform the decision making; surveillance to detect infected herds; control of the infection in in- fected herds and flocks; measures to prevent introduc- tion of a disease into free populations; certification of herds and flocks as a low-risk source. Objectives may range from preventing an increase in prevalence through to eradication. Excluded from this definition were: work done by an individual veterinarian on cases of disease on a farm; a private herd health program (also known as a private health and productivity scheme) offered by a vet- erinarian to an individual herd/farm.
Results
Countries and geographic regions
Out of a total of 109 invitees, 83 accepted nominations to collaborate in the project. Over 10 weeks between 26 February and 10 May 2018, 82 collaborators, working in- dividually or as teams within a country, completed the questionnaires. One participant died during the study so that data were obtained for 48 of the 49 countries that came together in the collaboration (Fig. 1): Argentina, Australia, Austria, Bangladesh, Belgium, Bhutan, Brazil, Canada, Chile, Colombia, Costa Rica, Czech Republic, Denmark, Ecuador, Finland, France, Germany, Greece, Iceland, India, Iran, Israel, Italy, Japan, South Korea, Lesotho, Mexico, Nepal, the Netherlands, New Zealand, Nigeria, Norway, Panama, Poland, Ireland, Slovenia, South Africa, Spain, Swaziland, Sweden, Switzerland, Thailand, United Kingdom, United States of America, Uruguay, Venezuela, Zambia, and Zimbabwe. There were six countries from Africa, 19 from Europe, two from the Middle East, seven from Asia, two from
Fig. 1Countries represented in this study
Australasia, three from North America, two from Cen- tral America and seven from South America. Informa- tion was not available from Russia and Central Asia as experts were either not identified or did not respond to invitations following nomination. These countries repre- sented a range of socioeconomic levels. Measured in terms of the United Nations Development Program (http://www.undp.org/content/undp/en/home.html) index ranking, they ranged from 1 (most developed) to 160 (least developed) out of 188 countries. Median ranking was 28.5 and the third quartile was 87.5, indicating that the participating countries were skewed towards those of higher socioeconomic status. More than one question- naire was completed for the following countries to enable region-specific information to be collated: Belgium (Wal- lonia, and whole country), Canada (Alberta, Atlantic Prov- inces, Ontario, and Québec), France (western France and whole country), Germany (Lower Saxony, Thuringia, and whole country), Spain (País Vasco, Castilla y Leon, Comu- nidad Valenciana, Galicia, and Navarra) and the United Kingdom (Great Britain, and Northern Ireland).
Within the broad geographic regions, the special inter- ests of the collaborators included field aspects of disease control, laboratory science, management and coordin- ation of disease control, communications and education, research and other skills related to disease control in- cluding public health. Collaborators from all 48 coun- tries had at least one special interest or skill area and those from 38 countries had special interests in two or more of these categories.
Animal populations at risk
All but one of 48 countries had at least two of the major types of farmed ruminants (dairy cattle, beef cattle, sheep, goats, camelids and deer), and 38 coun- tries (79%) had at least three (Table 1). Considering also other farmed ruminants (for example buffalo, bison), all 48 countries had at least three types or
species and two thirds of the countries (65%) had six or seven types/species.
About 70% of countries had more than 1 million cat- tle, sheep or goats and 20% of them had more than 10 million. Beef cattle and sheep were the most numerous of the livestock types; ten countries had > 10 million beef cattle while eight had more than 10 million sheep (Table 2). Goats were the second most numerous type of livestock; five countries had > 10 million. Three coun- tries had > 10 million dairy cattle whereas 20 countries had between 1 and 10 million. Dairy cattle, sheep and goats were present in all 48 countries, while beef cattle were not farmed in three countries (Bhutan, India, Nepal). Camelids and deer were the ruminants that were least often farmed, being reported in populations > 1000 from 12 countries and 15 countries, respectively. Coun- try level data are provided in Additional file 2: Table S1.
Median number of farms per country was highest for beef cattle (108,723), followed by sheep (36,313) and dairy cattle (21,090), whereas median herd sizes were highest for dairy cattle (108), sheep (68) and beef (50) . Box-and-whisker plots for the numbers of farms and herd sizes are presented in Additional file 3: Figure S1 to S4). These data are addressed further in the context of control programs for paratuberculosis in the section below.
Notifiability of paratuberculosis
Paratuberculosis was notifiable to the relevant authority in at least one of the farmed species in 35 (73%) of 48 countries, but the obligation to notify depended on the species affected (Table 3). Not all species were present in each country. It was notifiable regardless of species in Australia, Finland, Norway and Sweden and in all of the farmed species present in 32 countries. However, in eight countries (Argentina, Austria, Italy, Mexico, Slovenia, South Africa, United Kingdom, Venezuela) no- tification was not required for some species that were present, typically camelids and deer. In dairy cattle, it was notifiable in 35 of 48 (73%) countries. In beef cattle it was notifiable in 33 (73%) of the 45 countries in which this type of ruminant was farmed, in sheep and goats in 28 (64%) of 44 countries, in camelids in 12 of 28 (43%) countries and in farmed deer in 15 (50%) of 30 coun- tries. It was notifiable in buffalo in Colombia, Brazil, Italy, Japan and Switzerland and in bison in Canada and Switzerland.
Paratuberculosis was not notifiable in any species in 13 (27%) of 48 countries which were not clustered in any particular geographic region: Belgium, Czech Repub- lic, Denmark, Ecuador, France, Greece, India, Iran, the Netherlands, New Zealand, Nigeria, United States of America, and Uruguay (see Additional file 2: Table S2).
Table 1
Distribution of farmed ruminant types and species among 48 countries
No. types of ruminants
No. countries Dairy cattle, beef cattle, sheep, goats, camelids or deer
All types of farmed ruminants
1 1 –
2 9 –
3 7 1
4 3 9
5 15 7
6 13 17
7 – 14
Total 48 48
Under-reporting
Paratuberculosis was believed to be under-reported in 26 (74%) of the 35 countries in which it was notifiable, re- gardless of the geographic zone. The reasons given for under-reporting in decreasing order of frequency included lack of knowledge (farmer or veterinarian) or awareness of the clinical signs of disease (eight countries), reluctance to report due to farmer concerns about the consequences of quarantine or other action (5), lack of surveillance (4), no- tification only of clinical cases (3), poor diagnostic tests (2), lack of concern about paratuberculosis from govern- ment (2), lack of application of national legislation (1), government not notifying OIE (1), listing too recent for notification (1), deliberate choice of tests, the results of which are not notifiable (1), farmer culling of cases to avoid detection (1), and farmer fear of the stigma of iden- tification as a positive herd (1). There are also procedural causes of under-reporting, such as testing animals of rela- tively young age; this was cited by Italy where the product- ive life of dairy cows averages only 2.6 lactations.
Prevalence of paratuberculosis Herd-level prevalence
Notwithstanding missing information or unknown situations in some countries and even though a dis- tinction could not be made between apparent and true prevalence (Tables 4 and 5), paratuberculosis was
relatively common. Based on objective laboratory test data, herd-level prevalence < 1%, which may be con- sistent with definitions of freedom, was a feature in very few countries, for example 2 of 31 countries with prevalence data for dairy cattle and 2 of 15 countries with prevalence data for sheep (Table 4). But in 58%
of the 31 countries where there were prevalence esti- mates, > 20% of cattle, sheep or goat herds were in- fected. Similar proportions were observed where prevalence had been estimated from non-laboratory data (Table 4). Herd-level prevalence by country for each type of ruminant is presented in Additional file 2:
Table S3.
For dairy cattle there was a significant association be- tween herd size and herd-level prevalence of paratuber- culosis ( P = 0.001); for each one log increase in herd size the odds of a country having a higher category of prevalence increased by 9.7 (95% confidence limits 1.9–48.8). There were many missing data for either herd size or prevalence (20 countries had missing values for dairy cattle compared to 30 for beef cattle, 35 for sheep, 35 for goats and 45 for farmed deer;
camelids had very few observations).
Within-herd prevalence
Based on objective laboratory test data, within-herd prevalence in infected herds of < 1% was a feature in
Table 2Size of populations of each of the livestock types among 48 countries. Data are the number of countries
Population size Dairy cattle Beef cattle Sheep Goats Camelids Deer-farmed Other
< 1000 1 2 9 5 2
1000 to 10,000 1 1 1 8 6 8 3
10,000 to 100,000 2 2 7 14 1 5 2
100,000 to 1 million 20 13 15 7 7 2 6
1 million to 10 million 20 18 16 12 2
> 10 million 3 10 8 5 2
Unknown 2 1 9 9 5
Not applicable 3 16 19 26
Total 48 48 48 48 48 48 48
Table 3
Notifiability of paratuberculosis in each type of ruminant in 48 countries
Species/type
No. of countries % countries in which species is
applicable and paratuberculosis is notifiable
Notifiable Not notifiable Not applicable
Dairy cattle 35 13 72.9
Beef cattle 33 12 3 73.3
Sheep 28 16 4 63.6
Goats 28 16 4 63.6
Camelids 12 16 20 42.9
Deer-farmed 15 15 18 50.0
Other 10 11 27 47.6
very few countries, for example 1 of 32 countries with dairy cattle and 3 of 21 countries with beef cattle (Table 5). But within-herd prevalence > 10% for most species was common and > 15% was observed in at least 10% of countries. Similar proportions were observed where prevalence had been estimated from non- laboratory data (Table 5). Within-herd prevalence was often unknown (in 12 countries with dairy cattle, 18 with beef cattle, 26 with sheep and 27 with goats).
Within-herd prevalence by country for each type of ru- minant is presented in Additional file 2: Table S4.
Under-estimation of prevalence
Prevalence was believed to be underestimated in 29 (60%) of 48 countries regardless of the geographic zone.
Where reasons were given, in decreasing order of fre- quency they included: the type of tests used underesti- mate true prevalence (14 countries), lack of surveillance (12), lack of knowledge (farmer or veterinary) or aware- ness of the clinical signs of disease (2), reluctance to re- port due to farmer concerns about the consequences of quarantine or other action (1) and lack of concern about paratuberculosis from government (1).
Free-living ruminants and wildlife
Wildlife were known to have paratuberculosis in 18 (38%) of 48 countries which were not clustered in any particular geographic zone, but the situation was un- known in 26 countries (54%). Data were not available to separate infection with MAP from disease (pathology
associated with MAP infection) in these hosts. Often the hosts were known to be in contact with farmed live- stock. The list of species affected was extensive and in- cluded omnivorous, herbivorous and carnivorous mammals, including herbivorous marsupials, and even some birds as shown in Additional file 2: Table S5.
Control programs for paratuberculosis
A total of 22 (46%) of 48 countries had a control pro- gram for paratuberculosis in the period 2012 to 2018, i.e. one that commenced, was operating or ended in this period (Fig. 2) (Additional file 2: Table S6). Three of these countries had a prior program that ended before 2012, while 20 of them had a long-term program that will end after 2018. These long-term programs began as early as 1942 in the Netherlands and 1962 in Iceland, compared to as late as 2009 in New Zealand, with a me- dian year of 2000. Seven countries indicated that they will commence a new program in the period 2018–2020, including Chile and Slovenia, neither of which had had control programs before, and five countries which already had a control program in 2012–2018. Only two countries had a control program with a nominated end date, Australia (2018) and Canada (2018). Thus control programs began at different times and changed over time in each country. Major chronological events in the control programs in 22 countries are summarised in Additional file 2: Table S7.
While countries with a control program represented most geographic regions, more than half (14) were in
Table 4Herd-level paratuberculosis prevalence
aestimates in countries where laboratory testing had (or had not) been conducted.
Data are the number of countries at each prevalence level
Herd-level prevalence (%)a Dairy cattle Beef cattle Sheep Goats Camelids Deer-farmed
< 1 2 (1) 3 (3) 2 (1) 4 (2) 3 (1) 1 (1)
1–10 6 (2) 4 4 (2) 1 (1) 1
10–20 3 4 1 1 1
20–40 3 (1) 3 (1) 2 2
> 40 13 1 (1) 2 (1) 4 (1)
Total 27 (4) 15 (5) 11 (4) 12 (3) 4 (1) 2 (2)
aThe definition of prevalence and the way it was measured may have differed between countries and between species within countries
Table 5
Animal-level (within infected herds) prevalence
aestimates in countries where laboratory testing had (or had not) been conducted. Data are the number of countries at each prevalence level
Within-herd prevalence (%) Dairy cattle Beef cattle Sheep Goats Camelids Deer-farmed
< 1 1 3 (1) 1 (1) 2 1 (1) 1 (1)
1–5 5 6 (2) 2 2
5–10 10 5 2 2 (1) 1
10–15 7 (1) 2 (1) 1
> 15 3 2 2 3
Total 26 (1) 16 (3) 9 (2) 10 (1) 1 (1) 2 (1)
aThe definition of prevalence and the way it was measured may have differed between countries and between species within countries
Europe. The majority of countries without a control pro- gram for paratuberculosis were in South and Central America, Asia and Africa (19 of 25 countries) and only 5 (20%) were in Europe (Table 6).
Using the United Nations Development Program index ranking for 2015, countries without a paratuberculosis control program (median 73, range 19–160) were more disadvantaged than those that had one (median 14, range 1–119).
Number of farms or average herd size did not differ between countries with and without a paratuberculosis control program (Additional file 3: Figure S1-S4). How- ever, minimum herd size for camelids was larger in countries with a control program ( p = 0.046) and the maximum herd sizes for dairy cattle, beef cattle and sheep were larger for countries with a control program than those without a control program ( p = 0.04, p = 0.03 and p = 0.03, respectively).
Reasons for having a control program for paratuberculosis
All 22 countries with a control program for paratubercu- losis cited animal health as a reason for having a control program, whereas 90% cited reducing the economic or production losses (Table 7). Maintaining regional or international trade was a driver for 15 countries (68%) followed by animal welfare in 11 (50%). Public health was cited by eight countries (36%) in Asia (Japan, Korea and Thailand), Europe (Austria, Belgium, Germany and Republic of Ireland) and North America (Canada) as a reason for having a control program. An explicit state- ment regarding public health existed in the control pro- gram documentation or information of five countries (23%) (Austria, Germany, Japan, New Zealand and United Kingdom), two of which did not cite public health as a primary reason for having a control program.
The Republic of Ireland made reference to providing
“additional reassurance to the marketplace in relation to Ireland’s efforts to control Johne’ s disease” in the programme objectives (http://animalhealthireland.ie/
?page_id=340) while in the United Kingdom there was a general statement referencing the United Kingdom Food Standards Agency policy towards Johne ’ s disease and hu- man health, advising that the precautionary principle is
Fig. 2Countries represented in this study that had a control program for paratuberculosis between 2012 and 2018
Table 6
Number of countries with a paratuberculosis control program between 2012 and 2018 in each major geographic region
Geographic region No. of countries with a paratuberculosis control program between 2012 and 2018
No. countries in this study
Africa 1 6
Asia 3 7
Australasia 2 2
Europe 14 19
Middle East 0 2
North America 2 3
Central America 0 2
South America 0 7
Total 22 48
Table 7
Primary reasons for having a control program among 22 countries
Reason No. of
countries
% of countries
Animal health 22 100
Reducing production losses 20 90
Maintaining trade, regional or international 15 68
Animal welfare 11 50
Public health 8 36
adopted (https://www.afbini.gov.uk/articles/johnes-dis- ease).
Reasons for not having a control program for paratuberculosis
Of the 26 countries that did not have a program for paratuberculosis, nominated reasons were (Table 8): eco- nomic constraints, insufficient animal health resources beyond those already deployed to tackle other priority diseases, a lack of animal health capacity (including in- frastructure, operational resources, laboratory diagnostic services) and lack of feasibility due to inadequate control tools such as poor diagnostic tests and poor vaccines. In some countries paratuberculosis was deemed not preva- lent at herd or individual animal levels and was not con- sidered to be a problem relative to other animal health issues or it was present but was not considered to affect the animal population. The availability of paratuberculo- sis vaccine for use by farmers was cited as a reason for not requiring a control program by one country. Other reasons were given by ten countries including: paratu- berculosis is not perceived as an immediate problem;
there is lack of interest from farmers due to legislative restrictions; bureaucratic ignorance; cessation of funding;
lack of resources for collaboration between agencies and regulatory authorities; and lack of cost-benefit except for culling clinical cases.
Control programs for other diseases of livestock
Control programs for diseases other than paratuberculo- sis were present in many countries. Bovine tuberculosis (38 countries), bovine brucellosis (39), rinderpest (9) and foot and mouth disease (32) control programs are exam- ples, while programs for other diseases of livestock were present in 42 countries. Countries without a control pro- gram for paratuberculosis were neither more nor less likely to have a control program for one of the other dis- eases (Additional file 2: Table S8). Countries with a
control program for one of the other major livestock dis- eases were neither more nor less likely to have a control program for paratuberculosis.
Veterinary services on paratuberculosis for farmers beyond control programs
Services offered by veterinarians or others for individual farms included herd health programs in 27 of 48 coun- tries and advice on paratuberculosis for individual ani- mals in 32 of 48 countries. While the herd health services were available in a similar proportion of coun- tries with or without paratuberculosis control programs (14/22 and 13/26 countries, respectively; P = 0.34), ad- vice to farmers on individual cases of paratuberculosis was available in fewer countries without a control pro- gram for paratuberculosis than in countries with a con- trol program for paratuberculosis (12/26 and 20/22, respectively; P = 0.001) (Fig. 3).
Geographic organization of paratuberculosis control programs
It was very difficult to classify control programs strictly as regional or national. For example, in both Australia and Germany there was a national control program agreed through consensus of the regions, each of which had a State Government with jurisdictional control over biosecurity. In this case the programs were implemented and controlled by State jurisdictions, i.e. regionally, and they could differ from region to region in key aspects.
There were also differences in the species included in control programs between countries. The geographic organization of paratuberculosis control programs are summarised by country and species in Additional file 2:
Table S9.
Eighteen of 22 countries had a single national control program for the relevant species, while five had regional control programs (Belgium, Canada, Germany, France and Spain) that operated in some or most regions of the country, depending on the species. Two countries had
Table 8
Reasons stated for not having a control program for paratuberculosis among 26 countries
Reason No. countries % countries
Economic constraints 12 46
Animal health resources are currently deployed to tackle other priority diseases 11 42
Other 10 40
Lack of national/regional animal health capacity, infrastructure or operational resources 8 31 Paratuberculosis is not prevalent at herd or individual animal levels and is not considered to be a
problem relative to other animal health issues
8 31
Lack of laboratory diagnostic services 6 23
Paratuberculosis is present but is not considered to affect the animal population 5 19
Lack of feasibility due to inadequate control tools (eg. poor diagnostic tests, poor vaccines) 1 4 Paratuberculosis vaccine is available for use by farmers (there is no coordinated control program) 1 4
more than one national control program for dairy and/
or beef cattle while having a single national control pro- gram for other types of livestock (the Netherlands, United Kingdom). The regional program activities of re- gional or national programs were co-ordinated regionally in some countries (Australia, Belgium, Canada, France, Germany, Spain and United Kingdom) and nationally in all other countries. The control programs operating within species in different regions of a country were different from one another in some countries (Australia, Belgium, France, and Germany), but similar or identical to one an- other in other countries (Iceland, Korea, Spain, Thailand and the United Kingdom). In Canada, they were similar or identical in dairy cattle but different in beef cattle.
Objectives of control programs
The number of objectives specified by a country and the level of detail in each objective varied between countries and sometimes the methods of achieving an objective was expressed within the objective. This made assessing objec- tives difficult, and sometimes it was unclear whether objec- tives expressed were comprehensive statements or whether the descriptions were directly correlated with program activ- ities. Furthermore, objectives may have differed between the different types of livestock in a country. For example, two countries, Australia and the Netherlands, expressed their control program objectives differently for different livestock industries (dairy cattle, beef cattle and small ruminants).
Logically, program objectives and design reflected the stage of paratuberculosis control and initial prevalence.
For example, stamping out and then ongoing surveil- lance was the goal in countries where the disease was very uncommon or absent (such as Sweden and Norway) compared to culling cases in countries where the disease was common. For simplicity, the objectives were aggre- gated across types of livestock and are presented in
Table 9. The most common objective, expressed by 77%
of countries with a control program, was to reduce the prevalence of MAP infection, followed by the related ob- jective of reducing the incidence of clinical cases (45% of countries). The methods of doing so while also reducing spread to other herds was expressed quite specifically as a program objective by Spain: to reduce the prevalence of high faecal shedders or faecal shedding. Sweden, which was considered to be free of paratuberculosis, had the same objective but it would apply only if the disease were to be detected through surveillance. Reducing con- tamination of the human food chain (32% of countries), and the related objective (in two of these countries plus two others) of protecting markets for animal products was the next most common objective, meaning that 9 of 22 countries (41%) had an expressed objective related to product quality. Reducing spread to new farms or re- gions and certification or assurance of herds to provide a low risk source of stock including replacement stock, were each expressed by 27% of countries. Reducing eco- nomic losses and improving animal health, welfare and farm biosecurity were expressed by 18 and 9% of coun- tries, respectively. National or regional eradication of para- tuberculosis was an objective in only two countries (9%) (Norway and Sweden) while in three countries (14%) (Germany, the Netherlands and South Africa), eradication at herd level was an objective that could be chosen by a farmer and in one region of Belgium (Wallonia, with 120 herds participating out of 3000) regional eradication was a goal. Research on paratuberculosis and increasing educa- tion/awareness were not commonly expressed as objec- tives (9 and 4.5% of countries, respectively).
Performance indicators for control programs
Thirteen of 22 countries (56%) had performance indica- tors for their control program (Table 10). The most
Fig. 3Inclusion of advice on paratuberculosis in herd health programs, and availability of veterinary advice for individual animals in countries with and without a control program for paratuberculosis
common goal was an increase in the participation rate of herds in the control program (ten countries) followed by meeting targets in the number of low risk, free or cer- tified herds (5) and reductions in the number of infected animals or clinical cases detected (5). Norway, which had an eradication program, included active surveillance targets and completing farm-level post eradication checks as primary objectives. Sweden, which had ad- vanced beyond this to a surveillance program, used par- ticipation rate as the main performance indicator.
Review of control programs
The control programs of 13 countries had been subject to review at intervals ranging from annually (three
countries) to periodically every two to five years (four countries) and only once in the others. The one-time reviews were conducted as long ago as 2009 in the United States of America and as recently in 2015 in Germany. As a result of review there were substantial changes to the programs in four countries, changes to the objectives in three countries and changes to the operations and methods in nine countries. The pro- grams were not terminated in any country as a result of review.
Leadership of control programs
Leadership of the control programs among the 22 coun- tries was extremely variable, ranging from seven distinct components in Canada to a simpler structure with just one major nominated leader in 11 countries. In seven of these single-leader programs the leader was the govern- ment, while in the remaining four, leadership was pro- vided by a private organization, which in the case of Republic of Ireland was comprised of representatives of government and other sectors. There is likely to be some overlap in these classifications due to represen- tation of government in private organizations and vice versa, and in general the programs of most countries had diverse representation with 11 having more than one type of leadership component. Government was represented in at least 15 countries, farmer organiza- tions in at least 11, a veterinary organization in six, an industry milk organization in four, and an industry meat association in three.
Table 9
Objectives of paratuberculosis control programs among 22 countries
Objective No. countries % countries
Reduce prevalence of MAP (equivalent term = control) 17 77.3
Reduce incidence of clinical cases 10 45.5
Reduce MAP contamination in the human food chain/improve consumer safety 7 31.8
Provide confidence or assurance to, or safeguard markets (including export) 4 18.2
Reduce spread to new farms or regions 6 27.3
Certification of freedom or provide information on low risk herds as a source of replacement stock 6 27.3
Reduce production/economic losses 4 18.2
Risk management: determine risk in a herd; allow trade/marketing with an accredited or specified risk level; reduce within herd spread
4 18.2
Provide individual farmers with tools to manage JD (where manage = prevention or control) 3 13.6
Eradication of MAP from herds that aim to do this 3 13.6
Research including determination of prevalence or incidence 2 9.1
Elimination of high shedders or reduction of faecal shedding 2 9.1
Improve animal health and welfare and farm biosecurity 2 9.1
Country-level eradication (detect, control then eradicate) 2 9.1
Region-level eradication (detect, control then eradicate) 1 4.5
Education and awareness 1 4.5
Table 10
Performance indicators for paratuberculosis control programs in 13 countries
Performance indicator No. countries
Participation rate of herds 10
Number of low risk, free or certified herds 5 Incidence–the number of infected animals
or clinical cases detected
5
Prevalence - within herd 3
Number of infected herds 2
Number of MAP shedders 2
Number of herds in a test and cull plan 1
Animal production data 1
Active surveillance targets met 1
Program review and post eradication checks completed 1
Funding of programs
The sources of funding for control programs were di- verse and varied considerably between countries. Leader- ship was fully funded by government in 8 of 22 countries and partially funded by government in another eight countries. The next most common source was farmer organizations (eight countries) (see Additional file 2: Table S10). Program operations on the other hand were most likely to be funded wholly or partially by farmers (17 countries) or farmers in conjunction with government (eight countries) compared to government alone (three countries) (see Additional file 2: Table S11).
Participation and incentives
Participation in the control program was compulsory in at least some regions of nine countries (40%), and volun- tary in the other 13 countries. In Australia, participation was compulsory only in some states (Northern Territory, South Australia and Western Australia) while in the rest of the country it was voluntary except for notification of the disease. Compulsory participation was legislated in Austria, Japan, Norway and Switzerland and covered by ordinances in one region of Germany. Components of control that were compulsory varied from country to country according to the objectives of the program. For example, vaccination of sheep was mandatory in Iceland, where it was believed that without the legislation paratu- berculosis would be much more widespread. Active sur- veillance was mandatory in Japan. In South Africa, animals diagnosed with MAP had to be isolated and slaughtered under the supervision of a state official, in- contact animals had to be tested and infected herds/
flocks had to be placed under quarantine. Both Norway and Sweden had legislation that made reporting of any suspicion of MAP mandatory, and if MAP is detected (in any species), stamping-out, contact tracing, and eradication is compulsory.
Incentives for participation were identified in 15 coun- tries (68%), comprising ten out of the 13 (77%) with vol- untary programs and five out of the nine (56%) with compulsory programs ( p = 0.29). Incentives for participa- tion included higher prices or higher demand for ani- mals or animal products (four countries), avoidance of penalties such as market access restrictions for non- participation (ten countries), subsidised test costs (five countries) and various other incentives. In some coun- tries, the restrictions were onerous, as were the conse- quences of diagnosis which could mean exclusion from the market. For instance, in the Netherlands, dairy farmers are required by their milk processors to obtain a preferred herd status in a paratuberculosis control programme [53], meaning that herds have to be tested periodically and any test-positive cattle have to be culled.
Incentives and penalties in these countries are listed in Additional file 2: Table S12.
Financial assistance to farmers
At country level, full or partial financial support, assist- ance or compensation to farmers for one or more oper- ational aspects of the control program was provided in one or more species between 2012 and 2018 in 12 of the 22 countries (55%). Due to program review processes, this may have begun or ceased during this period, for ex- ample in Australia financial support was available to beef farmers until 2016, then ceased. Among the 12 countries in which financial support was provided, the compo- nents covered included the cost of testing (11 countries), the cost of culling infected livestock (seven countries), and the value of culled livestock (eight countries). All three components were covered in four countries, two were covered in six countries and one was covered in two countries.
Practices and tools used in control programs
Control programs for paratuberculosis in the 22 coun- tries were generally multi-component and involved many possible practices and tools (Table 11). Those employed depended on the objectives and hence also on the prevalence of paratuberculosis within the country.
The number of tools employed ranged from 2 to 11 among countries with 16 (73%) of the 22 countries employing five or more of the practices and tools listed in Table 11.
Table 11
Practices and tools used in control programs for paratuberculosis in 22 countries listed by frequency of inclusion
Tool No. of
countries
% of countries
Cull clinical cases 19 86.4
Hygienic rearing of neonates/juvenile livestock
17 77.3
Farm-level biosecurity to prevent introduction of infection
17 77.3
Test and cull subclinical cases 16 72.7
Environmental and pasture management 14 63.6
Communications program 14 63.6
Herd/flock assurance certification 13 59.1
Research program 11 50.0
Vaccination 7 31.8
Regional biosecurity to prevent introduction of infection
5 22.7
National biosecurity to prevent introduction of infection
4 18.2
Other 4 18.2
Stamping out infected herds/flocks 3 13.6
Individual animal assurance certification 3 13.6
